The rapid developments in recombinant DNA techniques have resulted in the identification and isolation of many novel genes, some of known function and some of unknown function. Invariably there is a need to express the gene in a heterologous cell system in order to produce material for structure-function studies, diagnostic reagents such as monoclonal or polyconal antibodies and material for in vivo activity testing and therapy.
Several alternative systems for the expression of foreign genes have been developed including systems based upon mammalian cells, insect cells, fungal cells, bacterial cells and transgenic animals or plants. The choice of expression system for a given gone depends upon the likely features of the encoded protein, for example any post-translational protein modifications needed for biological activity, as well as the objective of the study. Other important considerations for the investigator are the facilities available, time and cost involved in generating the amounts or recombinant protein required.
The most widely used and convenient system for the production of foreign proteins remains that based on the prokaryote Escherichia coli. The advantages of this system comprise the ease of gene manipulation, the availability of reagents including gene expression vectors, the ease of producing quantities of protein (up to a gramme in simple shake-flask culture), speed and the high adaptability of the system to express a wide variety of proteins.
Expression of any foreign gene in E. coli begins with the insertion of a cDNA copy of the gene into an expression vector. Many forms of expression vector are available. Such vectors usually comprise a plasmid origin of DNA replication, an antibiotic selectable marker and a promoter and transcriptional terminator separated by a mufti-cloning site (expression cassette) and a DNA sequence encoding a ribosome binding site. The method of transcriptional regulation varies between the various promoters now available (ptac, .lambda.pL, T7). The ptac and T7 expression based systems are controlled by the chemical inducer IPTG, whilst the .lambda. promoters are controlled by a temperature switch.
A problem encountered with E. coli based expression systems is the difficulty of producing material which is acceptable for therapeutic use. The use of complex media, antibiotic selection and potentially hazardous inducers such as IPTG may potentially render products such as recombinant antibody fragments produced by E. coli fermentation technology unacceptable to the regulatory authorities for clinical applications. Evidence demonstrating clearance of these agents from the final product must be provided in order to secure regulatory clearance. Clearance of these agents, and especially demonstrating such clearance, is expensive. It is therefore desirable that an expression system should avoid the three above-mentioned problems.
Avoidance of these problems is not straightforward. Plasmids, especially expression vectors, place a metabolic load on the host cell which acts as a selective pressure favouring loss of the plasmid from the cell. Therefore, in order to reduce the possibility of plasmid loss, or rearrangement of the plasmid to delete the expression activity, it is apparent that the metabolic load placed on a cell by other sources should be reduced to a minimum. Therefore, the use of complex media which contain, in addition to essential amino acids and minerals, a variety of naturally-sourced vitamins, cofactors and the like which alleviate the metabolic load on the cell is favoured.
The term "complex medium" is used herein according to its well-known signification in the art, that is to denote a medium the exact formulation and chemical composition of which has not been determined. Frequently, such media are at least partly derived from natural sources. For example, it is well known to include bovine serum preparations containing a variety of uncharacterised vitamins, growth factors and the like in cell culture media. In contrast, a defined medium is a medium which has been formulated from pure ingredients each of which is known, the medium therefore having a defined formula.
The use of a complex medium introduces the possibility of inclusion of a variety of agents derived from the natural source of certain medium components of which the regulatory authorities will require proof of clearance. Switching to the use of defined media would facilitate obtaining regulatory approval because it can be stated with certainty that the potentially harmful agents are absent from the cell culture. However, the efficiency of protein expression would also be reduced because the metabolic load on the cells in culture may be increased and plasmid expression systems may therefore be destabilised.
A manufacturer producing therapeutic products in cell culture is therefore faced with deciding which route will prove the most cost-effective. In the commercial production systems of the prior art, which use relatively unstable plasmid systems, the equation has often favoured the use of complex media and subsequent clearance demonstration.
The situation is similar in respect of antibiotics. These are commonly used to select against plasmid loss by inclusion of an antibiotic resistance gene on the plasmid. In the absence of antibiotic selection, the increased metabolic load placed on the cell by the plasmid acts to favour growth of plasmid free cells. Therefore, culture of transformed cells without antibiotic is not economic because, even though clearance of the antibiotic does not need to be demonstrated, the potential loss of plasmid from the cells may greatly reduce the efficiency of the expression system.
It should be borne in mind that it is not necessarily the cost of clearing the suspect agent from the product which is high, but rather the cost of providing evidence of this clearance. Thus, for example, IPTG may clear by degradation from the medium, but the cost of demonstrating the clearance of it and potential contaminants and degradation products thereof remains substantial.
It follows, therefore, that it would be desirable to be able to express proteins in a bacterial system in a defined medium in the absence of antibiotic selection and unacceptable inducers such as IPTG. However, due to the problems of plasmid instability, this aim has not been achieved in commercial expression systems of the prior art.
For example, a recent report describing expression of antibody fragments in E. coli host cells employs the antibiotic tetracycline to prevent plasmid loss from the cells (Carter et al. BioTechnology 10, 163-167, 1992). Carter et al employ an inducible expression system to prevent heterologous protein expression during host cell growth. Selective advantage for growth of plasmid free cells potentially enhanced by the use of defined medium is avoided by the use of tetracycline to select against such cells.
However, the use of antibiotic is required both during growth, evidently because the vector being used is not sufficiently stable, and during the expression phase. Although this system has the advantage that no inducer need be added, the requirement for antibiotic is not removed, as evidenced by the use of tetracycline in the expression system described. Therefore, the method of Carter et al will require the demonstration of clearance of antibiotic from the final product. This is hypothesised to be necessary because the expression vector used by Carter et al is not sufficiently stable to be cultured in the absence of antibiotic.
Antibodies and antibody fragments, especially recombinant or humanised derivatives thereof, are a class of proteins which it would be extremely desirable to be able to produce by recombinant DNA technology. By humanised antibodies, it is intended to refer to antibodies in which the constant regions are derived from human immunoglobulins, while at least the complementarity determining regions (CDRs) of the variable domains are derived from murine monoclonal immunoglobulins.
A number of improvements over natural immunoglobulins have been documented in the literature, which can only he put into practice by recombinant DNA technology. For instance, the production of CDR-grafted antibodies having CDRs from murine antibodies coupled to human framework regions can only be undertaken using a recombinant expression system. Furthermore, such systems are extremely useful for the production of antibody fragments which are not readily obtained by proteolytic cleavage, such as Fv fragments, and antibody fusions comprising an effector or reporter molecule attached to the antigen binding molecule.
Recombinant antibody fragments whether they be entire antibodies, Fab, Fab', F(ab').sub.2 or Fv fragments, consist of heavy and light chain dimers. A recombinant expression system should therefore be capable of expressing both heavy and light chain genes in such a manner as to render the individual peptides capable of self-assembly into the final product. This has been a stumbling block for recombinant antibody production, and indeed attempts have been made to solve the problem. An example of this is the production of "single chain" Fv fragments, wherein the heavy and light chain polypeptides are physically joined together by a flexible linker group. These molecules avoid the problems of chain association between free heavy and light chain polypeptides.
This system is not necessary, however, for the production of antibody fragments such as Fabs, which comprise heavy and light constant region chains as well as heavy and light variable region chains. For such applications it is desirable to express heavy and light chains separately in the same cell.
In order to facilitate correct assembly of heavy and light chains of antibody fragments, it is preferable to employ an expression system in which the chains are secreted into the culture medium rather than precipitated into the cell as inclusion bodies.
The use of E. coli signal sequences fused to polypeptides in order to facilitate their secretion by E. coli is known. However, it is apparent in a number of cases that secretion of heterologous proteins is even more deleterious for E. coli than accumulation of such proteins intracellularly. Accordingly, it has been found necessary to control the expression of heavy and light chain genes in order to achieve high cell growth and plasmid stability during the growth phase of a bacterial culture.